1. The Field of the Invention
The present invention relates to filtering multimedia content. More specifically, the present invention relates to methods, systems, and computer program products for automatically identifying and filtering portions of multimedia content during the decoding process.
2. The Prior State of the Art
Often, movies and other multimedia content contain scenes or language that are unsuitable for viewers of some ages. To help consumers determine whether a particular movie is appropriate for an audience of a given age, the Motion Picture Association of America (“MPAA”) has developed the now familiar NC-17/R/PG-13/PG/G rating system. Other organizations have developed similar rating systems for other types of multimedia content, such as television programming, computer software, video games, and music.
Both the quantity and context of potentially objectionable material are significant factors in assigning multimedia content a rating. However, a relatively small amount of mature-focused subject matter may be sufficient to remove multimedia content from a rating category recommended for younger children. For example, in a motion picture setting, a single scene of particularly explicit violence, sexuality, or language may require an “R” rating for what would otherwise be a “PG” or “PG-13” movie. As a result, even if an “R” rated motion picture has a general public appeal, individuals trying to avoid “R” rated content, and teenagers restricted by the “R” rating, may choose not to view a motion picture that they would otherwise desire to view if it were not for the inclusion of the explicit scene.
Many consumers may prefer an alternate version of the multimedia content, such as a version that has been modified to make the content more suitable for all ages. To provide modified versions of multimedia works, the prior art has focused on manipulating the multimedia source. The details of how multimedia content is modified depends largely on the type of access the source media supports. For linear access media, such as videotape or audiotape, undesired content is edited from the tape and the remaining ends are spliced back together. The process is repeated for each portion of undesired content the multimedia source contains. Due to the need for specialized tools and expertise, it is impractical for individual consumers to perform this type of editing. While third parties could perform this editing to modify content on a consumer's behalf, the process is highly inefficient because it requires physically handling and repeating the editing for each individual tape.
Modifying direct access media, such as DVD, also has focused on modifying the multimedia source. Unlike linear media, direct access media allows for accessing any arbitrary portion of the multimedia content in roughly the same amount of time as any other arbitrary portion of the multimedia content. Direct access media allows for the creation and distribution of multiple versions of multimedia content, including versions that may be suitable to most ages, and storing the versions on a single medium. The decoding process creates various continuous multimedia streams by identifying, selecting, retrieving and transmitting content segments from a number of available segments stored on the content source.
To help in explaining the prior art for creating multiple versions of a multimedia work on a single source, a high-level description of the basic components found in a system for presenting multimedia content may be useful. Typically, such systems include a multimedia source, a decoder, and an output device. The decoder is a translator between the format used to store or transmit the multimedia content and the format used for intermediate processing and ultimately presenting the multimedia content at the output device. For example, multimedia content may be encrypted to prevent piracy and compressed to conserve storage space or bandwidth. Prior to presentation, the multimedia content must be decrypted and/or uncompressed, operations usually performed by the decoder.
The prior art teaches creation and distribution of multiple versions of a direct access multimedia work on a single storage medium by breaking the multimedia content into various segments and including alternate interchangeable segments where appropriate. Each individually accessible segment is rated and labeled based on the content it contains, considering such factors as subject matter, context, and explicitness. One or more indexes of the segments are created for presenting each of the multiple versions of the multimedia content. For example, one index may reference segments that would be considered a “PG” version of the multimedia whereas another index may reference segments that would be considered an “R” version of the content. Alternatively, the segments themselves or a single index may include a rating that is compared to a rating selected by a user.
There are a variety of benefits to the prior art's indexing of interchangeable segments to provide for multiple versions of a multimedia work on a single storage medium. Use of storage space can be optimized because segments common to the multiple versions need only be stored once. Consumers may be given the option of setting their own level of tolerance for specific subject matter and the different multimedia versions may contain alternate segments with varying levels of explicitness. The inclusion of segment indexing on the content source also enables the seamless playback of selected segments (i.e., without gaps and pauses) when used in conjunction with a buffer. Seamless playback is achieved by providing the segment index on the content source, thus governing the selection and ordering of the interchangeable segments prior to the data entering the buffer.
The use of a buffer compensates for latency that may be experienced in reading from different physical areas of direct access media. While read mechanisms are moved from one disc location to another, no reading of the requested content from the direct access media occurs. This is a problem because, as a general rule, the playback rate for multimedia content exceeds the access rate by a fairly significant margin. For example, a playback rate of 30 frames per second is common for multimedia content. Therefore, a random access must take less than {fraction (1/30)}th of a second (approximately 33 milliseconds) or the random access will result in a pause during playback while the reading mechanism moves to the next start point. A typical 16×DVD drive for a personal computer, however, has an average access rate of approximately 95 milliseconds, nearly three times the 33 milliseconds allowed for seamless playback. Moreover, according to a standard of the National Television Standards Committee (“NTSC”), only 5 to 6 milliseconds are allowed between painting the last pixel of one frame and painting the first pixel of the next frame. Those of skill in the art will recognize that the above calculations are exemplary of the time constraints involved in reading multimedia content from direct access media for output to a PC or television, even though no time is allotted to decoding the multimedia content after it has been read, time that would need to be added to the access time for more precise latency calculations.
Once access occurs, DVD drives are capable of reading multimedia content from a DVD at a rate that exceeds the playback rate. To address access latency, the DVD specification teaches reading multimedia content into a track buffer. The track buffer size and amount of multimedia content that must be read into the track buffer depend on several factors, including the factors described above, such as access time, decoding time, playback rate, etc. When stored on a DVD, a segment index, as taught in the prior art, with corresponding navigation commands, identifies and orders the content segments to be read into the track buffer, enabling seamless playback of multiple version of the multimedia content. However, segment indexes that are external to the content source are unable to completely control the navigation commands within the initial segment identification/selection/retrieval process since external indexes can interact with position codes only available at the end of the decoding process. As a result, external segment indexes may be unable to use the DVD track buffer in addressing access latency as taught in the prior art.
As an alternative to buffering, segments from separate versions of multimedia content may be interlaced. This allows for essentially sequential reading of the media, with unwanted segments being read and discarded or skipped. The skips, however, represent relatively small movements of the read mechanism. Generally, small movements involve a much shorter access time than large movements and therefore introduce only minimal latency.
Nevertheless, the prior art for including multiple versions of a multimedia work on a single direct access media suffers from several practical limitations that prevent it from wide-spread use. One significant problem is that content producers must be willing to create and broadly distribute multiple versions of the multimedia work and accommodate any additional production efforts in organizing and labeling the content segments, including interchangeable segments, for use with the segment indexes or maps. The indexes, in combination with the corresponding segments, define a work and are stored directly on the source media at the time the media is produced. In short, while the prior art offers a tool for authoring multiple versions of a multimedia work, that tool is not useful in and of itself to consumers.
A further problem in the prior art is that existing encoding technologies must be licensed in order to integrate segment indexes on a direct access storage medium and decoding technologies must be licensed to create a decoder that uses the segment indexes on a multimedia work to seamlessly playback multiple versions stored on the direct access medium. In the case of DVD, the Motion Pictures Entertainment Group (“MPEG”) controls the compression technology for encoding and decoding multimedia files. Furthermore, because producers of multimedia content generally want to prevent unauthorized copies of their multimedia work, they also employ copy protection technologies. The most common copy protection technologies for DVD media are controlled by the DVD Copy Control Association (“DVD CCA”), which controls the licensing of their Content Scramble System technology (“CSS”). Decoder developers license the relevant MPEG and CSS technology under fairly strict agreements that dictate how the technology may be used. In short, the time and cost associated with licensing existing compression and copy protection technologies or developing proprietary compression and copy protection technologies may be significant costs, prohibitive to the wide-spread use of the prior art's segment indexing for providing multiple versions of a multimedia work on a single direct access storage medium.
Additionally, the teachings of the prior art do not provide a solution for filtering direct access multimedia content that has already been duplicated and distributed without regard to presenting the content in a manner that is more suitable for most ages. At the time of filing this patent application, over 5000 multimedia titles have been released on DVD without using the multiple version technology of the prior art to provide customers the ability to view and hear alternate versions of the content in a manner that is more suitable for most ages.
The prior art also has taught that audio portions of multimedia content may be identified and filtered during the decoding process by examining the closed caption information for the audio stream and muting the volume during segments of the stream that contain words matching with a predetermined set of words that are considered unsuitable for most ages. This art is limited in its application since it cannot identify and filter video segments and since it can only function with audio streams that contain closed captioning information. Furthermore, filtering audio content based on closed captioning information is imprecise due to poor synchronization between closed captioning information and the corresponding audio content.